Purpose: No evidence exists as to whether type 2 diabetes mellitus (T2DM) impairs clinical outcome from immune checkpoint inhibitors (ICI) in patients with solid tumors. Experimental Design: In a large cohort of ICI recipients treated at 21 institutions from June 2014 to June 2020, we studied whether patients on glucose-lowering medications (GLM) for T2DM had shorter overall survival (OS) and progression-free survival (PFS). We used targeted transcriptomics in a subset of patients to explore differences in the tumor microenvironment (TME) of patients with or without diabetes. Results: A total of 1,395 patients were included. Primary tumors included non–small cell lung cancer (NSCLC; 54.7%), melanoma (24.7%), renal cell (15.0%), and other carcinomas (5.6%). After multivariable analysis, patients on GLM (n = 226, 16.2%) displayed an increased risk of death [HR, 1.29; 95% confidence interval (CI),1.07–1.56] and disease progression/death (HR, 1.21; 95% CI, 1.03–1.43) independent of number of GLM received. We matched 92 metformin-exposed patients with 363 controls and 78 patients on other oral GLM or insulin with 299 control patients. Exposure to metformin, but not other GLM, was associated with an increased risk of death (HR, 1.53; 95% CI, 1.16–2.03) and disease progression/death (HR, 1.34; 95% CI, 1.04–1.72). Patients with T2DM with higher pretreatment glycemia had higher neutrophil-to-lymphocyte ratio (P = 0.04), while exploratory tumoral transcriptomic profiling in a subset of patients (n = 22) revealed differential regulation of innate and adaptive immune pathways in patients with T2DM. Conclusions: In this study, patients on GLM experienced worse outcomes from immunotherapy, independent of baseline features. Prospective studies are warranted to clarify the relative impact of metformin over a preexisting diagnosis of T2DM in influencing poorer outcomes in this population.

Current practice in cancer treatment collects multimodal data, such as radiology images, histopathology slides, genomics and clinical data. The importance of these data sources taken individually has fostered the recent rise of radiomics and pathomics, i.e., the extraction of quantitative features from radiology and histopathology images collected to predict clinical outcomes or guide clinical decisions using artificial intelligence algorithms. Nevertheless, how to combine them into a single multimodal framework is still an open issue. In this work, we develop a multimodal late fusion approach that combines hand-crafted features computed from radiomics, pathomics and clinical data to predict radiotherapy treatment outcomes for non-small-cell lung cancer patients. Within this context, we investigate eight different late fusion rules and two patient-wise aggregation rules leveraging the richness of information given by CT images, whole-slide scans and clinical data. The experiments in leave-one-patient-out cross-validation on an in-house cohort of 33 patients show that the proposed fusion-based multimodal paradigm, with an AUC equal to 90.9%, outperforms each unimodal approach, suggesting that data integration can advance precision medicine. The results also show that late fusion favourably compares against early fusion, another commonly used multimodal approach. As a further contribution, we explore the chance to use a deep learning framework against hand-crafted features. In our scenario characterised by different modalities and a limited amount of data, as it may happen in different areas of cancer research, the results show that the latter is still a viable and effective option for extracting relevant information with respect to the former.

Despite the advantages offered by personalized treatments, there is presently no way to predict response to chemoradiotherapy in patients with non-small cell lung cancer (NSCLC). In this exploratory study, we investigated the application of deep learning techniques to histological tissue slides (deep pathomics), with the aim of predicting the response to therapy in stage III NSCLC. We evaluated 35 digitalized tissue slides (biopsies or surgical specimens) obtained from patients with stage IIIA or IIIB NSCLC. Patients were classified as responders (12/35, 34.7%) or non-responders (23/35, 65.7%) based on the target volume reduction shown on weekly CT scans performed during chemoradiation treatment. Digital tissue slides were tested by five pre-trained convolutional neural networks (CNNs)—AlexNet, VGG, MobileNet, GoogLeNet, and ResNet—using a leave-two patient-out cross validation approach, and we evaluated the networks’ performances. GoogLeNet was globally found to be the best CNN, correctly classifying 8/12 responders and 10/11 non-responders. Moreover, Deep-Pathomics was found to be highly specific (TNr: 90.1) and quite sensitive (TPr: 0.75). Our data showed that AI could surpass the capabilities of all presently available diagnostic systems, supplying additional information beyond that currently obtainable in clinical practice. The ability to predict a patient’s response to treatment could guide the development of new and more effective therapeutic AI-based approaches and could therefore be considered an effective and innovative step forward in personalised medicine.

Background: Autologous hamstrings and patellar tendon have historically been considered the gold standard grafts for anterior cruciate ligament reconstruction (ACLR). In the last decades, the utilization of synthetic grafts has re-emerged due to advantageous lack of donor site morbidity and more rapid return to sport. The Ligament Augmentation and Reconstruction System (LARS) has demonstrated to be a valid and safe option for ACLR in the short term. However, recent studies have pointed out the notable frequency of associated complications, including synovitis, mechanical failure, and even chondrolysis requiring joint replacement.Case presentation: We report the case of a 23-year-old male who developed a serious foreign body reaction with wide osteolysis of both femoral and tibial tunnels following ACLR with LARS. During first-stage arthroscopy, we performed a debridement of the pseudocystic mass incorporating the anterior cruciate ligament (ACL) and extending towards the tunnels, which were filled with autologous anterior iliac crest bone graft chips. Histological analysis revealed the presence of chronic inflammation, fibrosis, and foreign body giant cells with synthetic fiber inclusions. Furthermore, physicochemical analysis showed signs of fiber depolymerization, increased crystallinity and formation of lipid peroxidation-derived aldehydes, which indicate mechanical aging and instability of the graft. After 8 months, revision surgery was performed and ACL revision surgery with autologous hamstrings was successfully carried out.Conclusions: The use of the LARS grafts for ACLR should be cautiously contemplated considering the high risk of complications and early failure.

In advanced or metastatic settings, Comprehensive Genomic Profiling (CGP) allows the evaluation of thousands of gene alterations with the goal of offering new opportunities for personalized treatment in solid tumors. This study evaluated the CGP Success Rate in a real-life cohort of 184 patients enrolled in a prospective clinical trial. CGP data were compared with the routine molecular testing strategy adopted in-house. Sample age, tumor area, and the percentage of tumor nuclei were recorded for CGP analysis. We found that 150/184 (81.5%) samples resulted in satisfying CGP reports. The CGP Success Rate was higher in samples from surgical specimens (96.7%) and in specimens that had been stored (sample age) for less than six months (89.4%). Among the inconclusive CGP reports, 7/34 (20.6%) were optimal samples, according to CGP sample requirements. Moreover, with the in-house molecular testing approach, we could obtain clinically relevant molecular data in 25/34 (73.5%) samples that had inconclusive CGP reports. In conclusion, despite the fact that CGP offers specific therapeutical options in selected patients, our data suggest that the standard molecular testing strategy should not be replaced in routine molecular profiling.

Recent years have witnessed the rise of pathomics as a mean to describe histopathological images with quantitative biomarkers for predictive and prognostic ends, combining digital pathology, omic science and artificial intelligence. This novel research branch is the counterpart of radiomics which pursues the same aims extracting knowledge from radiological images. In this paper, we present the design of a pathomic deep learning-based system to predict the treatment outcome in non-small cell lung cancer patients. We describe the system design and optimization under the condition of limited data and limited training, with corresponding tests. The experimental results show the feasibility of the proposed scalable architecture providing also a comparison between different transfer learning strategies.